Project Summary The dog, Canis lupus familiaris, has emerged as a powerful model system for analyzing the genetic basis of phenotypic diversity. Genomic tools coupled with the unique breed structure of dogs have permitted the association of over one hundred genetic regions with morphometric, behavioral, and disease phenotypes, many of which involve genes and phenotypes relevant to human health. During the past five years, studies in humans have identified copy-number and structural variation as important mutation types underlying disease susceptibility. In dogs, studies using microarrays have identified over 1000 sites of copy-number variation, but these studies are of poor resolution, have been limited to detecting large events, are unable to detect retroelement insertions, and are blind to balanced events such as inversions. In dogs and other model species, CNV studies to date have focused on modest numbers of samples (< 100) without detailed phenotype information. We propose to greatly extend our understanding of the distribution, biological basis, and phenotypic impact of dog CNVs through three specific aims designed to (1) define the breakpoint sequence of structurally variant haplotypes in a diverse panel of 41 dogs, (2) identify polymorphic canine genome sequences that are absent from the boxer-derived reference genome assembly, and (3) assess the frequency of common and rare copy-number variants using array intensity and genotype data from 4000 phenotyped individuals. This will be the first such study in a non-human mammalian species and will enable unique insights into the mutational processes underlying genomic stability. Furthermore, this research will directly contribute to future and ongoing effort to map medically relevant phenotypes in dogs using array-based genotyping, strengthening these efforts aimed at identifying and understanding pathways that underlie human disease.